Practical Example of Jury Selection

Let's assume for a practical example a trial that is set in Port Orchard, Kitsap County, Washington State. It's a civil trial in which the plaintiff, a single mother of two and former decade-long employee of a mid-sized Kitsap company, is suing the company for gender discrimination.

Plaintiff's counsel elects to contract the predictioneering services of Jury Games. After agreeing to terms, the Jury Games team learns from plaintiff's counsel its strategy for the case: it will produce evidence, in the form of employee reviews, that the plaintiff typically performed within the top 10% of all employees for the company during her tenure, and yet she was passed over for promotion by less qualified men. The legal team already knows instinctively that it will likely want to favor women over men on the jury, favor employees rather than business owners, and favor younger jury members rather than older.

Statistical Profiling for Value Indexing

Jury Games will then start conducting statistical profiling of the local area population, which in this case would mean profiling the population of the Kitsap peninsula. Game Theory From this research, we'd likely learn about various different demographic groups that will naturally cluster into profile segments. One could be a blue-collar employee of the ship yard. Another might be a mid-level Navy enlisted man. Another could be a recently retired man in his late 60s. And yet another might be a businessman from Bainbridge Island.

Using this data and counsel's trial strategy, Jury Games constructs a set of value indexes to measure, including, but not limited to, the following: desire for prestige and respect; empathy for others; success in promoting one's own will; intercommunication need; sense of fair play versus a sense of uniform justice or equality; and philosophy of balance ratio between free market capitalism and individual liberty and prosperity.

After checking-in with the legal team to verify and validate our value indexes and assumptions, Jury Games builds a set of juror profiles based on mapping our local survey data to our value index matrix. We then generate a predictioneering model that conforms with our research, and then preload it with our profile value index data.

Just prior to voir dire, once plaintiff's counsel has received juror questionnaire data from the court, Jury Games uses this data to map each candidate to a pre-established profile. At that point, Jury Games might need to make minor customizations to the data of each candidate profile based on outlier juror questionnaire data.

Voir Dire Process

During voir dire, as the legal team asks questions of the jury panel, Jury Games alters and refines each candidate's specific data for each value index, depending on an individual juror's response. To use a simple example, let's say counsel asks the panel, "Have any of you been in a situation where you felt your boss was being unfair to you?" Given how a number of candidates might respond, counsel might follow-up with, "How did that make you feel?" One candidate might respond by saying she felt very hurt and angry over the injustice, which means that Jury Games would likely inflate the number associated with her value index pertaining to justice and possibly empathy.

After the conclusion of voir dire and the commencement of peremptory challenges, the laptop screen shifts from that of data input and adjustment to predictioneering view. The predictioneering system then runs multiple simulations on each juror candidate, and all possible juries are then constructed based on the inclusion of a particular juror on the jury. The system then reports a numerical value between 0 and 100 indicating to what level a juror candidate should be on the jury. Let's say in this case, the system reported a value of 61 for one candidate, which is somewhat positive. Plaintiff's counsel, however, may have received a "bad vibe" from juror candidate 1, so despite the 61 score, Plaintiff exercises a peremptory challenge on this candidate. For another candidate, the system reports a value of 76. Plaintiff's counsel wants this juror; however, the defense ultimately exercises a peremptory challenge and the juror is dismissed. Each time a candidate is accepted or rejected, the predictioneering simulations are all rerun given the confirmation of who is and could be on the jury. This system then continues through the exercise of peremptory challenges and/or the acceptance of the panel.